Logging when API calls fail lead users to believe the experiment failed

[clayburn]

At the end of the Build Validation Script executions, some API calls are made to populate the summary. These calls can fail for a few reasons (e.g. lack of API authorization, networking issues). If they do fail, the user is presented with some cryptic errors in eye-catching red text:

Response status code: 0
ERROR: Unable to connect to server to fetch build scan https://develocity.example.com/s/foo
Response status code: 0
ERROR: Unable to connect to server to fetch build scan https://develocity.example.com/s/bar

Often, the user will then click the build scan links and see that they do in fact exist.

In these cases, when a user is running the script manually, the experiment will still provide users with the quick links for investigation, so the experiment was not entirely a failure. The lack of summary should be treated more like a warning. It would be good to be more helpful and let the user know what happened, what the impact was, and how to remediate it.